January 27, 2016 at 4:42 am
Hi Folks,
I just want to raise this question in this forum after a discussion in our development group today.
Our BI architect is encouraging us to use a OLAP cube for the reports, almost as a must, because of the grouping that the report datasets needed.
I understand that using a cube everything is pre aggregated (of course if it is well designed) and the performance for reports or dashboard with a high granularity should be better than aggregating the data on-the-fly using SQL. However, some report developers complains about the skills required.
For me the skills are not so critical, or at least not the reason to have a performance penalty. I see other possible drawbacks like cube processing time, what happened if the cube processing crashes, the time to develop a cube, etc.
The BI Arch. said that a Tabular Cube is not so complicated to develop, so that is our first approach.
What is your opinion or experience in this topic, could you please share it with me?
Any comment will be appreciates,
Kind Regards,
January 27, 2016 at 5:35 am
If your OLAP is well designed then my recommendation is for it. It doesn't touch the OLTP database(s) i.e. doesn't put extra locking on the tables. OLAP can be also used by reporting services.
However, the mind-set thinking is different from the one needed for T-SQL development.
Igor Micev,My blog: www.igormicev.com
Viewing 2 posts - 1 through 1 (of 1 total)
You must be logged in to reply to this topic. Login to reply